蒜种颗粒离散元模型参数标定

Parameters Calibration of Discrete Element Model for Garlic Particles

为确定蒜种离散元模型仿真参数,采用多球聚合的方法建立了蒜种离散元模型,并对可实现测量的模型参数(蒜种-钢板恢复系数、蒜种-蒜种恢复系数、蒜种-钢板静摩擦因数、蒜种-蒜种静摩擦因数)进行了物理试验测定。对不易测量的模型参数(蒜种-钢板滚动摩擦系数、蒜种-蒜种滚动摩擦系数)进行了仿真试验标定,应用、建立了两种不易测量的接触参数与蒜种堆休止角、密相区域圆半径的二次回归模型,最终求解寻优得到参数最优值。通过上述试验方法,得到蒜种离散元模型参数:蒜种-钢板恢复系数为0.511,蒜种-蒜种恢复系数为0.487,蒜种-钢板静摩擦因数为0.473,蒜种-蒜种静摩擦因数为0.503,蒜种-钢板滚动摩擦因数为0.111,蒜种-蒜种滚动摩擦因数为0.108。最后,在得到的接触参数下进行仿真验证试验,结果表明:蒜种堆休止角、密相区域圆半径的仿真值与实际值的相对误差分别为1.36%、1.50%,无明显差异,排种功率、排种速度的仿真与试验值在整体变化趋势上具有一致性,验证了蒜种离散元模型与仿真试验的有效性,可为大蒜播种机械仿真设计与优化提供参考。

In order to determine the parameters of garlic particles in discrete element model simulation,this study established garlic discrete element model by the method of multi-sphere aggregation,The model parameters which can be measured(garlic-steel coefficient of restitution,garlic-garlic coefficient of restitution,garlic-steel coefficient of static friction,garlic-garlic coefficient of static friction)were measured by physical test,The model parameters which are not easy to measure(garlic-steel coefficient of rolling friction,garlic-garlic coefficient of rolling friction)were calibrated by simulation test,The quadratic regression models of two hard to measure model parameters,repose angle of garlic pile and dense phase region radius were established by central composite design,and the optimal value of parameters is obtained.The discrete element model parameters of garlic species were obtained by the above experimental methods:the garlic-steel coefficient of restitution is 0.511,the garlic-garlic coefficient of restitution is 0.487,the garlic-steel coefficient of static friction is 0.473,the garlic-garlic coefficient of static friction is 0.503,the garlic-steel coefficient of rolling friction is 0.111,the garlic-garlic coefficient of rolling friction is 0.108.Finally,the simulation test is carried out under the contact parameters obtained,the results show that,the relative errors of the simulated and actual values of the repose angle and the dense phase region radius of garlic pile were 1.36%and 1.50%respectively,with no significant differences,and the simulation and test values of planting power and planting speed are consistent with the overall variation trend,which verifies the effectiveness of the discrete element model and simulation test of garlic seeds,and provides a reference for the simulation design and optimization of garlic seeding machinery.